Mixture Ratio Design Optimization of Coal Gangue-Based Geopolymer Concrete Based on Modified Gravitational Search Algorithm

A green concrete, new type of coal gangue-based geopolymer concrete, was prepared. Coal gangue geopolymer concrete contains many mineral admixtures and alkaline activators; the concrete mixture ratio design has always been a complex problem. The framework of the mix design optimization by the proposed method is established in this work. The paper aims to minimize the economic cost under the premise of ensuring the strength and workability of coal gangue-based geopolymer concrete. Gravitational search algorithm (GSA) has the advantages of faster convergence speed and stronger exploitation performance compared with the traditional optimization algorithms. However, GSA tends to premature convergence and local optimum, with weak search ability. Therefore, chaotic map is introduced in the work here. Gravitational search algorithm was modified based on Chebyshev map in chaotic theory, and the modified equations were derived. The modified algorithm was verified by the calculation of typical functions. And results from traditional GSA and GSA modified by another chaotic mapping, logistic mapping, were compared and the characteristics of different GSA were analyzed and concluded. After that, the mix design of geopolymer concrete based on coal gangue with different strength grades was optimized with the modified GSA. Through analysis of the optimization results, cost variation of different strength grade coal gangue-based geopolymer concrete was revealed. Costs declined significantly; the higher the grades within a certain strength range, the more saved. Therefore, it can be inferred that the modified gravity search method provides a reliable tool for the optimization of mixture ratio of similar geopolymer concrete.